LSTM + FC training stagnation

[shubhvachher]

check the outputs of the LSTM… See what it is predicting to see why the accuracies don’t change.

I’m facing something similar here MQRNN (Seq2Seq LSTM + MLP) model gives constant predictions